Dramatic changes in composition and metabolism of glycosphingolipids (GSLs) have been observed in essentially all experimental and human cancers, resulting in formation of a variety of tumor-associated antigens defined by monoclonal antibodies, although the functional significance of tumor-associated GSL antigens is not known. On the other hand, a number of concurrent studies have suggested two distinct functional roles of GSLs, i.e., (1) GSLs as adhesion molecules based on GSL-GSL interactions, and synergistic effect of this system with other cell adhesion systems based on integrin receptors or leccams. Tumor-associated GSLs could be the targets of leccam recognition. . (2) GSLs and their metabolites as modulators of transmembrane signaling. E.g., GM3, together with lyso-PC, lyso-GM3, and N,N-dimethylsphingosine (derived from sphingosine), were found to be strong functional modulators of key molecules playing roles in transmembrane signaling (e.g., protein kinase C, EGF receptor kinase). They may also be modulators of functional units (e.g., "oncoproteins") which regulate cell proliferation. Studies along lines 1 and 2 above are expected to provide new means of correcting or blocking tumor cell proliferation, invasiveness, and metastasis, e.g., application of GSLs, their modified structures, or MAbs directed to them for inhibiting interaction of tumor cells with capillary endothelial cells of specific organs, or blocking interaction of tumor cells with cells of neighboring tissues (this direction of study could be called "anti-adhesion therapy"). We may also discover ways to block transmembrane signaling mechanisms which are apparently enhanced in a variety of tumor cells. These mechanisms can be suppressed by employing physiological inhibitors (e.g., metabolites of GSLs, phospholipids, and SPN), with consequent preferential blocking of tumor cell proliferation. A preliminary experiment indicated that this "anti-signaling" approach indeed inhibited tumor cell malignancy. Aberrant expression of A antigen (deletion, overexpression of A antigen in A tumors, or expression of incompatible A in O or B tumors) is a topic of long-standing interest in clinical immunology, in view of the close association between aberrant A antigen expression and patient prognosis. Since we have cloned the A and B genes, in-depth understanding of the molecular genetic basis of aberrant ABO status in human cancer is another specific goal of this proposal.